Expert information. Types of genetically engineered animals

Transcription

1 Expert information Committee for Genetics and Breeding of Laboratory Animals Types of genetically engineered animals Status: June 2017 Authors: Ingrid Renner-Müller, Munich Johannes Schenkel, Heidelberg Bernhard Aigner, Munich

2 Exclusion of liability The use of the publications and statements of GV-SOLAS and the implementation of the information contained therein are explicitly at your own risk. GV-SOLAS and the authors cannot be held responsible for any accidents or damage of any kind arising from the use of the publication (e.g. due to lack of safety information), irrespective of their legal grounds. Liability claims against GV-SOLAS and the authors for any damage of material or immaterial nature caused by the use or non-use of the information or the use of incorrect and/or incomplete information are generally excluded. Legal and damage claims are therefore excluded. The publication including all contents has been compiled with the greatest care. However, GV-SOLAS and the authors assume no liability for the topicality, correctness, completeness and quality of the information provided. Printing errors and false information cannot be completely excluded. GV-SOLAS and the authors do not assume any liability for the topicality, correctness and completeness of the contents of the publication, as well as for printing errors. GV SOLAS and the authors cannot assume any legal responsibility or liability in any nature for any incorrect information and the resulting consequences. Only the owners of the websites printed in these publications are responsible for the contents of these Internet pages. GV-SOLAS and the authors therefore explicitly disassociate themselves from all third-party contents. Liable in accordance with the German press laws: the Board of Directors of GV-SOLAS. Contents 1 Introduction Definitions Experimental purpose General steps of the generation of mutant or transgenic animals 4 2 Non-transgenic animals harbouring genetically engineered, defined mutations 4 3 Genetically engineered, transgenic animals Transgenes Non-homologous recombination of transgene and host genome Additive gene transfer Use of CRISPR-Cas9 as a transgene Suppression of specific host gene products Trap constructs Transposons Homologous recombination of transgene and host genome Control of the transgene function Induction of transgenes Conditional mutagenesis Techniques for the generation of transgenic animals Random transgene integration by non-homologous recombination DNA microinjection Viral vectors Sperm-mediated gene transfer Defined, enzyme-mediated transgene integration by non-homologous or homologous recombination Defined, non-enzyme-mediated transgene integration by homologous recombination Embryonic stem cells (ES cells) Somatic cell nuclear transfer (cloning) 19 4 Selected literature General literature Species-specific publications Technical developments 21 page 2 of 22

3 1 Introduction Generation and breeding of mutant animals is done by breeding of animals exhibiting spontaneous mutations or by inducing experimental alterations in the genome. This can be carried out by unspecific mutagenesis using conventional methods (e.g., chemical or physical mutagenesis) or much more precisely by using genetic engineering techniques. The induction of genetically engineered alterations which are stably transmitted to the offspring can result in mutations of defined endogenous DNA sequences in non-transgenic animals (see section 2) or - by integration of experimentally transferred gene constructs - in the generation of transgenic animals (see section 3). Genetically modified lines must be systematically analyzed on the occurrence for pain, suffering and harm of the animals. Genetically engineered animals are used for the analysis of gene function and gene regulation (functional genome analysis), for the generation of disease models as well as for the production of biologically active proteins or modified animal products (gene farming). Generation of a high number of different genetically engineered lines is feasible for each of the ca. 20,000 protein coding mammalian genes and the additional regulatory genome loci. Novel animal models are established by the generation of genetically engineered animals using two complementary approaches: the usually applied gene-based approach (reverse genetics) or the phenotype-based approach (forward genetics). The gene-based approach (reverse genetics) is carried out by modifying the function of a defined genome sequence chosen for a specific experimental purpose and by subsequently examining the consequences of the experimental alterations on the phenotype. Predictions on the specific phenotypic outcome are limited. For the complementary phenotype-based approach (forward genetics), transgenic animals are produced by random insertional mutagenesis via non-homologous DNA recombination and are subsequently analyzed to detect a defined phenotype chosen for a specific experimental purpose. Genetically engineered animals exhibiting the mutant phenotype of choice are further bred, and subsequently the host genome is analyzed to detect the causative mutation. Usually a loss of function is present in the host genome of these mutant lines. Data about the numbers of animals used for experiments in Germany are published annually by the Ministry of Nutrition and Agriculture (BMEL) and show that far more than two out of three vertebrates used in animal experiments are mice; half of them are genetically modified animals. In addition, zebrafish is also used in higher numbers as genetically modified vertebrates in animal experiments ( The legal classification of non-transgenic animals harbouring genetically engineered mutations (see section 2) has not yet been implemented in the respective German regulations and is not within the scope of this manuscript. The induction of genetically engineered alterations in animals without germ-line transmission of the induced mutation to the offspring also is not within the scope of this manuscript. 1.1 Definitions Using enzyme-mediated techniques like the application of sequence-specific nucleases (designer nucleases) can result in the induction of genome alterations (genome editing) without the integration of foreign DNA into the host genome. Animals derived from those experiments are genetically engineered, non-transgenic animals (see section 2). Transgenic animals are experimentally generated mutants by transfer and integration of transgenes into the host genome. Animals exhibiting the stable genome alteration in the germ cells (gem-cell gene transfer) are further used to breed transgenic lines. Transgenesis of the animals is proven by identifying the transgene in the host genome. The function of the transgene is demonstrated by detection of transgene expression (transcribed mrna or protein) and/or transgene-specific alterations of the host phenotype. Simply detectable marker genes (reporter genes) are used for various research purposes. The choice of species and strain for the production of genetically engineered animals depends on the experimental purpose and comprises the classical laboratory mammals (mouse, rat, rabbit) as well as farm animals (pig, ruminants) and further vertebrates (birds, amphibians, fish). Published mutant mouse lines are collected in internet databases, where, however, only a page 3 of 22

4 part of the published lines derived from the additive gene transfer as the type of induction of the mutation is covered ( Production of valid data in such experimental projects generally requires the generation and analysis of several independent genetically engineered lines, as unintended genetic and/or epigenetic alterations due to the techniques used for the genetic modification may occur. If these unintended alterations are not linked to the mutant locus of choice, they can be removed by conventional breeding. The use of inbred strains allows the analysis of the influence of the genetic background on the phenotype of the genetically engineered line by the generation of congenic strains. This is carried out by backcrossing the mutant or transgenic locus from the original strain (donor) into the genome of a novel strain (recipient) for ten subsequent generations which results in the exchange of the genetic background in the newly established strain except of a ca. 20 cm long genomic fragment (covering ca. 1% of the genome) harbouring the mutant or transgene locus. The international nomenclature for genetically modified lines is described in the internet ( 1.2 Experimental purpose The experimental purpose involving genetically engineered animals determines the strategy and the technique used for the genetic modification of the respective animal models. The experimentally induced mutation may result in an additional function by over- and/or ectopic expression of a transgene (gain of function) and/or in the partial or complete loss of function of host genes (loss of function). This includes the inactivation of specific genomic sequences (knockout), defined genomic modifications (knockin), suppression of the synthesis of specific gene products (knockdown, gene silencing) as well as random mutagenesis of the host genome (insertional mutagenesis). The purpose of using genetically engineered, non-transgenic animals (see section 2) usually is the partial or complete loss of function of defined host genes chosen for the respective research projects. The aim of the project may require transgene activity in specific cell types/tissues and/or at defined life stages (spatio-temporal control). Controllable transgenes may be reversibly or irreversibly functionally activated or inactivated. 1.3 General steps of the generation of mutant or transgenic animals The generation of mutant or transgenic animals involves the use of assisted reproduction techniques (ART) as biotechnological methods which may cause unintended epigenetic alterations in the founder animals leading to phenotypic alterations. The process generally consists of the following steps: - Production of a sequence-specific nuclease system or a functional transgene - Transfer of the sequence-specific nuclease system or the transgene into host cells of donor animals. Depending on the technique used, gametes or progenitor cells thereof, early embryonic stages, embryonic stem cells or somatic cells are used. - Generation of embryos derived from the treated cells and/or embryo transfer in pseudopregnant recipient animals (foster mothers) - Analysis of the offspring of the recipients (founders = F0 animals) on the mutation or integration of the transgene in the host genome - Breeding of genetically engineered lines with the genetically engineered founder animals - Genotypic and phenotypic characterization of the genetically engineered lines 2 Non-transgenic animals harbouring genetically engineered, defined mutations Enzyme-mediated genetic engineering techniques may induce alterations on defined genomic locations (site-specific recombination) of embryos without the integration of foreign DNA into the host genome. The purpose of using genetically engineered, non-transgenic animals page 4 of 22

5 usually is the partial or complete loss of function of defined host genes chosen for the respective research projects. Therefore, artificial, programmable endonuclease systems harbouring components for binding specific DNA sequences like CRISPR-Cas9 (clustered regularly interspaced short palindromic repeat associated protein 9; or class 2 CRISPR endonuclease systems with analogous function), transcription activator-like effector nucleases (TALEN) or zincfinger nucleases (ZFN) are constructed which bind to selected host genome sequences and cause double strand breaks at these sites. The CRISPR-Cas9 system consists of two parts, the enzyme (Cas9) and short project- and sequence-specific nucleic acids, that makes it superior to both other systems in view of the application of the technique. The subsequent DNA repair via the mechanism of non-homologous end joining (NHEJ) results in mutant animals harbouring alterations with a length of usually few base pairs (indels) at the site of the double strand break. Site and type of the mutations are not exactly controllable leading to different mutations in the affected animals. Control of the induction of mutations is increased using refined versions of the CRISPR-Cas9 system. Simultaneous mutation of multiple defined genomic sites in a given embryo has also been done. Standard technique The generation of the mutant animals usually is carried out using the microinjection into zygotes and contains following steps: - Production of the programmable, sequence-specific nuclease system (CRISPR-Cas9, TALEN, ZFN). The CRISPR-Cas9 system is superior to both other systems in view of the application of the technique. - Hormonal super-ovulation of donor animals - Mating to fertile males - Breeding control - Preparation of zygotes from the oviduct - In vitro injection of the sequence-specific nuclease system into the pronuclei and/or in the cytoplasm of zygotes (RNA molecules or ribonucleoprotein (RNP) complexes but not DNA constructs are usually applied for to avoid unintended random integration of DNA fragments after unspecific degradation.). Alternatively, electroporation is used as transfer technique. - Transfer of the embryos into the oviduct of the pseudo-pregnant recipients (foster mothers) - Analysis of the offspring of the recipients for mutations in the selected genomic site. Animals exhibiting suitable mutations are subsequently chosen for further breeding. - Breeding of mutant lines - Genotypic and phenotypic characterization of the mutant lines Genotype of animals A considerable number of the offspring of the recipients (founders = F0 animals) may exhibit mutations in the selected gene sequence usually comprising few base pairs (indels). The mutant F0 animals can be classified in two groups: (1) Animals exhibiting the identical mutation in all cells of the organism, giving rise to heterozygous mutants, homozygous mutants, or alterations of both alleles by different mutations (bi-allelic mutation, compound heterozygosity). (2) Mosaic animals exhibit the mutation not in all, but only in a fraction of the cells. After mating the mutant founders to wild-type animals, heterozygous mutant and non-mutant offspring appear in the subsequent F1 generation if the experimentally induced mutation is present in the germ cells. Presence of more than one mutant allele for the selected gene sequence in the germ cells of the mated founder animals may result in mutant F1 offspring harbouring different mutations. In this context, the appearance of more than two different mutations in the germ cells of founder animals has already been described. After mating heterozygous mutant F1 offspring harbouring the identical mutation, homozygous mutant animals appear in the F2 generation according to the Mendelian rules of inheritance. The use of inbred strains as genetic background leads to genetically uniform homozygous mutant animals within a given line. For obtaining valid data, generation and analysis of several independent lines is also recommended in projects with genetically engineered, non-transgenic animals. Off-target effects page 5 of 22

6 of the enzyme system used like triggering unintended double strand breaks leading to mutations at non-selected sites of the host genome or unintended integration of foreign DNA are described. Use of nucleases with binding sites within some distance to each other may result in the deletion of the genome fragments between the selected binding sites. 3 Genetically engineered, transgenic animals For the generation of transgenic animals, a research project-specific DNA construct (transgene) is generated in vitro and transferred into the host genome by various techniques. Defined enzyme-mediated transgene integration after co-injection of transgene and sequence-specific nuclease system (CRISPR-Cas9 or class 2 CRISPR endonuclease systems with analogous function, TALEN, ZFN) also allows the generation of mutations in the host genome where when compared to spontaneous mutations - the use of transgenes is no longer detectable in the genetically engineered, mutant animals by DNA sequencing techniques (see section 3.2.2). 3.1 Transgenes The transgene constructed in vitro with the use of cloning vectors and/or by applying artificial DNA synthesis is composed of genomic DNA or recombinant DNA. Bacterial cloning vector sequences are removed before the use of the transgene. The transgene may contain isogenic, allogenic and/or xenogenic DNA sequences including functional prokaryotic genome sequences referring to the chosen host organism. The size of the transgenes usually ranges between few kb and several 100 kb. The intact integration into the host genome has to be analyzed especially when using long constructs. The performance of the transgene may be modified by neighbouring endogenous sequences and therefore strongly depends on the integration site of the host genome (position effect). Random or defined integration of the transgene into the host genome can be achieved by applying various transfer techniques. The defined integration via homologous recombination requires a specific type of transgene, the targeting vector (see section 3.1.2). For random transgene integration into the host genome, all established transfer techniques (see section 3.2) can be used. On the other hand, the defined integration of the transgene into the host genome usually is conducted by enzyme-mediated techniques using sequence-specific nuclease systems (see section 3.2.2) and/or by non-enzyme-mediated techniques using embryonic stem cells (see section ) or somatic cell nuclear transfer (see section ). Transgenes may be tested in suitable in vitro systems before their use in animals. However, the results may not reliably represent their performance in in vivo systems Non-homologous recombination of transgene and host genome Non-homologous recombination of the transgene and the host genome result in a random integration of the transgene into the host genome; only the enzyme-mediated transgene integration also enables a non-homologous recombination event at a defined site of the host genome (site-specific recombination; see section 3.2.2). Transgene integration into the host genome by non-homologous DNA recombination usually is carried out for the production of transgene-specific RNA and/or transgene-specific proteins (additive gene transfer). The transgene may contain one or multiple genes chosen for specific research purposes with the identical or different spatio-temporal expression. In addition, the CRISPR-Cas9 system may be used as a transgene by itself (see section ). Trap constructs are applied for random mutagenesis of the host genome (see section ). Transposons are used for the random, enzyme-mediated gene transfer (see section ) Additive gene transfer Transgenes for the additive gene transfer consist of three basic domains: (1) regulatory domain, (2) coding region and (3) termination unit. (1) The regulatory domain determines the site page 6 of 22

7 (ubiquitous or cell type-specific) and the time point/period (constitutive or at defined life stages; spatio-temporal expression; inducible or non-inducible; see section 3.1.3) as well as the strength of the transgene expression. Complete regulatory domains including cis-regulatory elements (specific promoter, binding sites of transcription factors) and trans-regulatory (long-range) elements (enhancer, insulator, locus control region [LCR], repressor/silencer) induce the optimal activity. Optimization of the transgene design covers the regulatory control of the transgene expression. (2) The coding region contains the chosen gene. Non-coding sequences including regulatory sequences and an exon/intron structure promote the effective expression. Alternatively, non-protein-coding RNA may be expressed (see section , gene silencing). (3) The termination unit provokes the proper expression of the transgene. If an endogenous host gene is used for the additive gene transfer, discrimination of the expression of transgene and endogenous gene on the protein level may be done by applying different alleles or by linking additional short translated DNA sequences to the transgene which give rise to additional epitopes (peptide tags) to be used for detection by specific antibodies. However, they may also interfere with the protein activity. Expression cassettes include tested regulatory and termination units as well as intermediate DNA sequences (multiple cloning site [MCS]) for the insertion of the coding sequences of choice. Using these tools, the expression of the transgene is predictable at some degree. Sequences exhibiting shielding properties (insulator, matrix attachment region [MAR], scaffold attachment region [SAR]) against adjacent host genome sequences may increase specificity and copy number-dependence of the transgene expression, however, contrasting results are published after the application of these tools. Autonomously replicating transgenes without integration into the host genome (like mammalian artificial chromosomes [MAC]) result to date in an instable inheritance of the foreign DNA. Inactivation or ablation of defined cells or cell compartments may be induced by spatiotemporal expression of toxic products or receptors of toxic substances as transgene (e.g., the diphtheria toxin or the receptor of the diphtheria toxin) Use of CRISPR-Cas9 as a transgene The CRISPR-Cas9 system (or class 2 CRISPR endonuclease systems with analogous function) may be used as a DNA construct by itself for the additive gene transfer to evoke spatiotemporal alterations in the host genome of the transgenic animals. The CRISPR-Cas9 system consists of two parts, the enzyme (Cas9) and short research project- and sequence-specific nucleic acids. Thus, both parts or the enzyme Cas9 alone may be applied as a transgene. The latter case requires the exogenous application of sequence-specific nucleic acids. Application of DNA constructs coding for an active Cas9 enzyme usually aims for the generation of short indels in selected genomic sequences. Alternatively, use of DNA constructs coding for enzymatically non-active Cas9 proteins in combination with additional effectors linked to the enzymatically nonactive Cas9 proteins may lead to sequence-specific activation or deactivation of the expression of selected genomic sequences. Spatio-temporal control of the transgene system is feasible by using appropriate promoter sequences and/or controllable systems (see section 3.1.3). Off-target effects of these systems may occur. Transgenes harbouring the coding sequences of sequence-specific nucleases (and potentially further genes, so-called cargo genes) flanked by parts of its own target sequence may induce sequence-specific double strand breaks by the expressed nuclease at the site of the complete target sequence in the host genome of the produced transgenic organisms. This increases the chance of homologous recombination events at this site leading to the subsequent insertion of the transgene at the site of the double strand break. Transfer of the transgene via self-duplication in diploid cells may directly induce homozygous transgenic genotypes of the cells. Thus, introduction of this system into a population may yield sequence-specific genetic modifications in the organisms of the population ( mutagenic chain reaction, gene drive). page 7 of 22

8 Suppression of specific host gene products RNA interference (RNAi) by additive transfer of DNA sequences expressing non-proteincoding regulatory RNA sequences (ncrna) is applied for the targeted post-transcriptional suppression of the synthesis of defined host genome products (gene knockdown, gene silencing). Therefore, the transgene codes for the expression of naturally occurring or artificially designed RNA molecules that specifically inactivate complementary mrna molecules of the host cell by producing RNA double strand molecules together with the mrnas. This leads to the subsequent degradation of these structures thereby decreasing or completely inhibiting the production of the respective host protein. Specificity and level of inactivation vary in the publications of such projects, and the complete inactivation turned out to be difficult Trap constructs Trap constructs contain regulatory elements in various versions (enhancer-, promoter-, gene-, poly (A)-trap) and marker genes (reporter gene). They are used for the induction of random insertional mutations via non-homologous DNA recombination in the host genome, typically in embryonic stem cells. In addition, retroviral vectors are used as trap constructs. The vectors may induce extensive alterations at the insertion site of the host genome. The mutant cells are clonally cultivated, analyzed for the successful expression of the marker gene as well as for the induced mutation caused by the integration of the trap construct into the host genome. After generating transgenic animals with the mutant cells of choice, in vivo analysis of the mutated host genome sequences (loss of function) is carried out. The mutant cells generated by this technique in various approaches worldwide are listed in several online databases ( Transposons Use of two-piece DNA transposon systems derived from various species like e.g., sleeping beauty (SB) or piggybac (PB) gives rise to a random, enzyme-mediated gene transfer. The transgene is flanked on both sides by the target sequence (inverted terminal repeats, ITR) of the enzyme transposase. Transgene integration at random sites induces only small alterations of the adjacent host genome. Usually DNA microinjection of transgene and transposase with temporally limited activity is carried out for the additive gene transfer that may result in increased efficiency of transgene integration, insertion of single copies per integration site and multiple segregating integration sites in the host genome when compared to the conventional technique (see section ). The length of the transgenes used is limited for the successful in vivo application. In addition, DNA transposons may be applied for the induction of random insertional mutations in the host genome of mammals which includes the generation of two transgenic lines. The first transgenic line harbours the transgene flanked by the target sequence of the transposase, and the second line is transgenic for the enzyme transposase. Breeding both lines may result in transgenic offspring exhibiting multiple random insertional mutations for the subsequent use in phenotype-based forward genetics approaches. Published work described that complete reconstitution of the wild-type allele is possible by the enzyme-mediated removal of integrated transgenes of the piggybac (PB) system through a subsequent transposition Homologous recombination of transgene and host genome Targeted integration of the transgene into the host genome allows the induction of complete inactivation (knockout) or defined modifications (knockin) of specific host genes. It is conducted by enzyme-mediated techniques with the use of sequence-specific nuclease systems (see section 3.2.2), and is carried out directly in embryos or indirectly in embryonic stem cells (see section ) or in cells that are subsequently used for the somatic cell nuclear transfer (see section ). Usually double stranded or single stranded DNA targeting vectors are applied as transgenes. Due to the high efficiency of the sequence-specific nuclease systems, page 8 of 22

9 targeting vectors without marker genes for the selection process may be used thereby allowing the generation of mutations in the host genome where when compared to spontaneous mutations - the use of transgenes is no longer detectable in the genetically engineered, mutant animals by DNA sequencing techniques (see section 3.2.2). Alternatively, the conventional, non-enzyme-mediated targeted integration of the transgene into the host genome is feasible in vitro by homologous recombination in embryonic stem cells (see section ) or with the somatic cell nuclear transfer (see section ). This method obligatorily requires a targeting vector as specific transgene and results in a low efficiency. In the targeting process, known genome sequences are replaced with the transgene leading to the defined mutation of endogenous DNA sequences. Thereby, homologous recombination of DNA sequences identical to the host genome on both sides of the transgene with the endogenous host sequences occurs (gene targeting). The length of the introduced mutation is highly variable covering point mutations as well as long DNA fragments. Subsequently, transgenic animals are generated using these in vitro specifically mutated cells. Prerequisite for the construction of targeting vectors is the knowledge of the DNA sequence and the structure of the chosen host gene ( Integration sitespecific expression cassettes (see section ) are available as targeting vectors allowing the controlled production of transgene-specific proteins (additive gene transfer) by the integration of the transgene at a defined position of the host genome like e.g., the ROSA26 locus. Both ends of the classical targeting vector harbour DNA sequences which are usually a few kb in length, homologous to the chosen integration site of the host genome and derived from isogenic DNA of cells planned to be used for the generation of the specific mutation. Double stranded or single stranded targeting vectors with only short homologous DNA sequences work well when combined with the use of sequence-specific nuclease systems. The intended mutation is inserted between both homologous ends. The targeting vector also includes marker genes which are essential for the selection of the cell clones with the desired alteration. The positive marker gene usually remains in the genome of the established transgenic animals, but may be also removed from the host genome in vitro or in vivo in the case that appropriate transgene constructs are used Control of the transgene function Transgene activity is primarily controlled transcriptionally by using appropriate regulatory elements. This does not modify the integrated transgene by itself. Furthermore, the defined modification of the transgene locus in the transgenic animals is feasible by using conditional mutagenesis. Both systems can be combined Induction of transgenes On the transcription level, the spatio-temporal transgene expression can be reversibly regulated. Inducible transgenes with appropriate promoter sequences allow the functional activation or inactivation of the transgene. The passive state of the transgene shows no or only small activity, whereas reversible induction increases the activity. Inactivation of the transgene function is conducted in the reverse direction. In both cases, for practical reasons, the inactivation of the inducible promoter should be almost complete. Promoter control is conducted (1) by application of exogenous substances. Endogenous substances combined with their target promoters do usually not give rise to a stringent regulation of the induction of transgene expression. Triggering the transgene induction by synthetic substances or extrinsic proteins may avoid additional unintended effects of the organism or unintended endogenous cross reactions. Recently, controlled application of essential amino acids via food combined with the use of optimized promotor elements (amino acid response elements) has been described to result in a low basal expression and a high inducible and reversible expression of transgenes. Experiments for the spatio-temporal, reversible induction of transgene expression via exposure to blue light combined with synthetic transactivator and promoter elements (LightOn system) have principally been successful, but depend on the penetration of the light into the body tissues. page 9 of 22

10 On the other hand, promoter control is conducted (2) by two-piece transgene systems. This requires the generation of two transgenic lines each of them harbouring one part of the system as a transgene. After breeding double transgenic animals by mating both transgenic lines, the spatio-temporal control of the expression of the first transgene is conducted by tissuespecific and/or time point-specific promoter sequences of the second transgene which drive the expression of a product that subsequently interferes with the promoter of the first transgene thereby resulting in its activity or inactivity. (3) The two-piece transgene system can be combined with exogenously supplied substances. Often systems using the application of tetracycline or derivates thereof (e.g., doxycycline) are administered. Here, the expression of the chosen transgene is based on the transgenic expression of the tetracycline-dependent, active transactivator tta (tet-off system) or the tetracycline-dependent, inactive reverse transactivator rtta (tet-on system) under tissue-specific promoter control. The tet-off system relies on the expression of the chosen transgene via binding of the usually tissue-specifically expressed active transactivator tta to the promoter element TRE (tetracycline responsive element) of the chosen transgene. Application (via food, drinking water or injection) of tetracycline or doxycycline leads to the binding of this exogenous factor to the transactivator tta which subsequently prevents the binding of the now inactive transactivator tta to the promoter element TRE and thereby prevents the expression of the chosen transgene. The reverse tet-on system relies on the absence of the binding of the usually tissuespecifically expressed, reverse transactivator rtta to the promoter element TRE. Thus, no expression of the chosen transgene takes place. Application of tetracycline or doxycycline leads to the binding of the newly forming rtta-doxycycline complex to the promoter element TRE, and subsequently to the induction of the expression of the chosen transgene Conditional mutagenesis The spatio-temporal function of genes can be analyzed with inducible systems allowing the activation or inactivation of transgenes (conditional mutagenesis). The induction gives rise to a gene knockout/-in and is irreversible. For this purpose, two-piece recombination systems like the bacterial cre/loxp system (cre = causes recombination, loxp = locus of crossover (X) of bacteriophage P1) are applied. The cre recombinase catalyzes the recombination of the DNA fragment between the two short identical loxp recognition sites. Therefore, two transgenic lines are generated: the first line exhibits the usually cell type-specific expression of the cre recombinase as a transgene, and the second line exhibits the chosen gene that is flanked on each side with the loxp sequences (floxed target gene). The transgenic lines are generated by using enzyme-mediated, sequence-specific nuclease systems and/or by non-enzyme-mediated techniques via targeting vectors. Breeding both lines yields double transgenic animals where the recombination is triggered by the cre recombinase. Tissue- and/or time point-specific promoter sequences of the transgene harbouring the cre recombinase control the spatio-temporal regulation of the conditional mutagenesis. The spatio-temporally controlled and/or incomplete conditional mutagenesis results in the generation of mosaic animals. Prolonged and/or high expression of recombinases may induce unintended interactions with endogenous sites of the host genome as erroneous target sites (off-target effects) and the phenotypic consequences resulting thereof. This may be overcome by using heterozygous cre recombinase-transgenic animals. Moreover, use of the cre-transgenic animals as an additional negative control group is recommended in such experiments. Published cre recombinase-transgenic mouse lines are collected in online databases (e.g., Position and orientation of the loxp recognition sites determine the arising recombination event. Position of the loxp recognition sites on one DNA molecule and identical orientation give rise to the deletion of the intermediate DNA fragment. Position of the loxp recognition sites on two DNA molecules (e.g., on an integrated transgene and an exogenously supplied transgene; or on two integrated transgenes at different genome loci) and identical orientation give rise to the translocation of both DNA molecules at these sites. Position on one DNA molecule and inverse orientation result in the cre-mediated inversion of the intermediate DNA fragment which may induce a change in the function of the corresponding DNA fragment. Possible applications of the recombination systems are the generation of genome recombinations with a length of up to page 10 of 22

11 several 100 kb (deletion, inversion, translocation), the removal of marker genes, gene knockin applications, or the generation of defined point mutations in the host genome. For instance, stop sequences flanked by loxp sequences can be inserted in front of specific gene sequences, thereby transcriptionally or translationally preventing the expression of the respective gene. The cre recombinase removes the stop sequence and irreversibly gives rise to the expression of the gene. The application of the induction approach (see section ) on the cre/loxp system can be done by using the respective inducible promoter sequences for the expression of the cre recombinase. Thus, transcription of the transgenic cre recombinase can be reversibly controlled by combining the cre/loxp system with the tet-on system or the tet-off system and the exogenous application of tetracycline or doxycycline in a three-piece transgene system: Transgene 1 harbouring the tissue-specific promoter for the expression of the transactivator, transgene 2 with the promoter element TRE (tetracycline responsive element) for the expression of the cre recombinase and transgene 3 with the loxp recognition sequences of the cre recombinase. This requires the breeding of triple transgenic animals (see section ). The genome recombination caused by the transgenic cre recombinase is irreversible. The reversible control of the cre function on the posttranslational level is done by linking the transgenic cre recombinase to a mutant ligand binding domain of the estrogen receptor. Endogenous estrogen does not bind to the mutant ligand binding domain. Without the exogenous application of the estrogen derivate tamoxifen, the transgenic fusion protein of cre recombinase and estrogen receptor cannot enter through the cell membrane to the nucleus as the site of its function. This occurs only after exogenous application of tamoxifen which subsequently binds to the fusion protein. The genome recombination caused by the transgenic cre recombinase is irreversible. Soybeans contain phytoestrogens which may interfere with the estrogen receptor and cause unintended reactions. Therefore, food free of or containing only low levels of phytoestrogens should be used. The same is true for bisphenol A that is found especially in the plastic of water bottles. Conditional mutagenesis is also induced by similar tools like the yeast-derived flp/frt system and further yeast-derived systems, the bacterial dre/rox system, mutant systems of the cre/loxp- and flp/frt system, and the bacteriophage-derived ΦC31/attP/attB system. These recombinases can be also combined for the efficient integration of transgenes at target sites. Simultaneous use of two different systems with non-interacting recognition sequences and with the recognition sequence located both in the host genome and on an exogenously applied transgene construct may cause in vitro or in vivo the exchange of the DNA fragment between the different recognition sequences via the activity of the two respective recombinases (recombinase-mediated cassette exchange, RMCE). Simultaneous use of two different systems with non-interacting recognition sequences and with inverse orientation of the respective recognition sites on a transgene may cause a first recombination in form of an inversion (and usually the loss of function of the intermediate DNA fragment), and a second recombination in form of the repeated inversion (and the rescue of the function of the intermediate DNA fragment). 3.2 Techniques for the generation of transgenic animals The function of the transgene potentially depends on the integration site in the host genome (position effect). Non-enzyme-mediated direct transfer methods include DNA microinjection, viral vectors and sperm-mediated gene transfer that result in the random insertion of the transgene into the host genome (non-homologous DNA recombination, see section 3.2.1). Enzyme-mediated transfer methods by using sequence-specific nuclease systems are much more efficient and allow the targeted insertion of the transgene into the host genome (see section 3.2.2). Alternatively, targeted insertion of the transgene into the host genome is possible indirectly by non-enzyme-mediated, in vitro operating tools via homologous DNA recombination in embryonic stem cells (ES cells) or in cells subsequently used for the somatic cell nuclear transfer (cloning), however this is much less efficient (see section 3.2.3). Transgenic animals harbouring the genetic modification in one allele after nonhomologous DNA recombination are named as hemizygous transgenic animals. Transgenic animals harbouring the genetic modification in one allele after homologous DNA recombination page 11 of 22

12 are named as heterozygous transgenic animals. Transgenic animals harbouring the genetic modification in both alleles are named as homozygous transgenic animals irrespective of the kind of DNA recombination. Compared to the low efficiency of the basal DNA microinjection technique, the establishment of novel techniques for the generation of transgenic animals may have the potential to lower the number of animals needed in the experiment Random transgene integration by non-homologous recombination Non-homologous recombination of transgene and host genome without the use of sequence-specific nuclease systems results in the random insertion of the transgene into the host genome. Transgenes for the additive gene transfer are suitable tools for this techniques (see section 3.1.1) DNA microinjection Microinjection of DNA constructs is the basal standard method of the additive gene transfer for all vertebrate species. The efficiency of the method, i.e. the number of transgenic offspring related to the number of microinjected embryos, without the use of sequence-specific nuclease systems is described to be 1-5 % for laboratory animals. Standard technique The generation of transgenic animals via DNA microinjection contains following steps: - Production and in vitro testing of the transgene construct - Hormonal super-ovulation of donor animals - Mating to fertile males - Breeding control - Preparation of zygotes from the oviduct - In vitro injection of the transgene into the pronuclei of zygotes - Transfer of the embryos into the oviduct of the pseudo-pregnant recipients (foster mothers) (Alternative: Prolonged in vitro incubation of the embryos (blastocysts) and trans-cervical transfer into pseudo-pregnant recipients) - Analysis of the offspring of the recipients for transgenesis - Parallel breeding of several transgenic founder lines - Genotypic and phenotypic characterization of the transgenic lines Genotype of animals Up to 25% of the offspring of the recipients exhibit the random insertion of few or up to several hundred transgene copies usually in one non-segregating site of the host genome. The insertion site may contain complete and fragmented transgene copies in both orientations mixed with endogenous host sequences. The host genome may be altered at the boundaries to the transgene by recombinations with a length of several kb. The transgenic animals (founders = F0 animals) can be classified in two groups, while both kinds of mutations may also appear combined in a given founder animal: (1) Hemizygous transgenic animals exhibit the transgene at the identical genomic site in all cells of the organism, (2) while mosaic animals exhibit the transgene not in all, but only in a fraction of the cells. After mating the founders to non-transgenic wild-type animals, hemizygous transgenic and nontransgenic offspring occur in the subsequent F1 generation if the experimentally induced mutation is present in the germ cells. Founder animals harbouring segregating transgene insertion sites may give rise to the production of independent transgenic lines. After mating hemizygous transgenic F1 littermates, homozygous transgenic animals appear in the F2 generation according to the Mendelian rules of inheritance. The use of inbred strains as genetic background leads to genetically uniform homozygous transgenic animals within a given line. Usually transgene locus and transgene expression remain stable after the establishment of the first few generations of a transgenic line. page 12 of 22

13 Insertion site and transgene locus of the transgenic lines are random that may give rise to potential differences in time point, site and strength of the transgene expression in the established lines of a given research project. This makes it necessary to generate and analyze several independent lines for the project. Using methods leading to the random insertion of the transgene into the host genome results in the unintended alteration of host genes (insertional mutation) in a low number of the established transgenic animals. This may give rise to unforeseen, not transgene-specific alterations of the physiological phenotype of the usually affected homozygous transgenic animals (in the case of a recessive mutation). In the event that the insertional mutation is lethal, no hemizygous transgenic and/or homozygous transgenic animals appear further on from the affected life stage Viral vectors Various viral vectors are used for the insertion of transgenes into the host genome. Generation of transgenic vertebrates is carried out with genetically modified viral particles. Using their own enzymes, retroviruses complementarily transcribe their RNA genome into double stranded DNA that is subsequently inserted into the host genome. Transgenic lines showing stable integration of the gene construct into the host genome and function of the transgene are usually produced with recombinant lentiviral vectors. Translocation of the lentiviral genome into the host cell nuclei takes place via nucleopores which allows the direct transgene insertion. Translocation of other retroviral genomes into the host cell nucleus requires the division of the respective cell. Lentiviral vectors offer the possibility of the virus-independent, spatio-temporal control of the transgene by using suitable promoter sequences. The injection technique used causes a less strong physical manipulation of the embryos when compared to the conventional DNA microinjection method. Standard technique The generation of transgenic animals with lentiviral vectors contains following steps: - Production of the DNA transgene construct - Production of infectious lentiviral particles which contain the transgene in the virus genome in form of RNA sequences, by using packaging constructs and specific cell lines - Hormonal super-ovulation of donor animals - Mating to fertile males - Breeding control - Preparation of zygotes from the oviduct - Infection of the zygotes (or alternatively of non-fertilized oocytes and subsequent in vitro fertilization [IVF]) with the recombinant retroviral particles by injection of the lentiviral particles into the perivitelline space or by co-incubation of the embryos after removal of the zona pellucida with the lentiviral particles - Transfer of the infected embryos into the oviduct of the pseudo-pregnant recipients (foster mothers) - Analysis of the offspring of the recipients for transgenesis - Parallel breeding of several transgenic lines - Genotypic and phenotypic characterization of the transgenic lines Genotype of animals Up to a high percentage of the offspring of the recipients may be transgenic, and transgene copies may be inserted in several segregating sites of the host genome of a founder animal. Thus, production of independent transgenic lines may be feasible from a given founder animal. Each of the random insertion sites harbours one transgene copy and exhibits alterations of the host genome on the boundaries to the transgene with a length of only few nucleotides. The transgenic animals (founders = F0 animals) can be classified in two groups: (1) Hemizygous transgenic animals exhibit the transgene at the identical genomic site in all cells of the organism, (2) while mosaic animals exhibit the transgene not in all, but only in a fraction of page 13 of 22

14 the cells. After mating the founders to non-transgenic wild-type animals, hemizygous transgenic and non-transgenic offspring appear in the subsequent F1 generation if the experimentally induced mutation is present in the germ cells. After mating hemizygous transgenic F1 littermates, homozygous transgenic animals appear in the F2 generation according to the Mendelian rules of inheritance. The use of inbred strains as genetic background leads to genetically uniform homozygous transgenic animals within a given line. The insertion site of the transgenic lines is random that may give rise to potential differences in time point, site and strength of the transgene expression in the established lines of a given research project. This makes it necessary to generate and analyze several independent lines for the project. Using methods leading to the random insertion of the transgene into the host genome results in the unintended alteration of host genes in a low number of the established transgenic animals (insertional mutation, see section ). Disadvantages of the system are increased demands on the biosecurity level for the work with infectious lentiviral particles, the limited length of the transgene and a potential loss of function of integrated transgenes in the following generations via endogenous methylation of the transgene locus. Afterwards, the specific methylation status of the transgene locus and the transgene expression level of the produced lines seem to be stable during the breeding of further generations Sperm-mediated gene transfer The sperm-mediated gene transfer is a simple technique to generate transgenic animals. However, up to now published results demonstrate the principal successful use in mammals but considerable difficulties in establishing the technique and achieving reproducible results. In total, up to now especially the successful incubation of sperm and transgene is not implemented as a routine method. Several methods and combinations of methods are described for the sperm-mediated gene transfer, and various methods are published for the incubation of sperm or its progenitor cells and the transgene. (1) In vitro incubation of sperm and transgene; subsequently applying in vitro fertilization (IVF): - Preparation of sperm from fertile donor animals - In vitro incubation of sperm and transgene - Preparation of oocytes after super-ovulation of donor animals - In vitro fertilization (IVF) of non-fertilized oocytes with the treated sperm (in special cases by intracytoplasmic injection of the treated sperm [ICSI] into the non-fertilized oocytes) - Transfer of the zygotes or higher developmental stages after in vitro culture into pseudopregnant recipients (foster mothers) (2) In vitro incubation of sperm and transgene; subsequently applying direct artificial insemination (AI): - Preparation of sperm from fertile donor animals - In vitro incubation of sperm and transgene - Artificial insemination of recipients with the treated sperm (3) Injection of the transgene into the testes: - Direct injection of the transgene into the testes of fertile animals with or without subsequent electroporation (alternative: use of viral vectors) - Mating of the treated recipients to untreated females (4) In vitro incubation of sperm progenitor cells and transgene: - Preparation of sperm progenitor cells from the testes of fertile donor animals - In vitro incubation of sperm progenitor cells and transgene - Injection of the transgenic sperm progenitor cells into the testes of unfertile recipients - Mating of the recipients to untreated females Using these methods may give rise to offspring that may be transgenic up to a high percentage. The transgenic animals exhibit the random insertion of transgene copies in one or more segregating site(s) of the host genome. Little is known about these transgene loci. page 14 of 22